DE69516181T2 - RUTHENIUM-RHENIUM / COAL CATALYST FOR HYDROGENATION IN AQUEOUS SOLUTION - Google Patents

Description

BACKGROUND OF THE INVENTION 1. Field of the invention

The invention relates to an improved catalyst and a process for using the catalyst for hydrogenation in aqueous solution. More particularly, but not by way of limitation, the invention relates to the production of tetrahydrofuran, γ-butyrolactone, 1,4-butanediol and the like from a hydrogenatable precursor such as maleic acid, succinic acid or the like in aqueous solution in the presence of hydrogen and a catalyst consisting essentially of highly dispersed reduced ruthenium and rhenium on a carbon support.

2. Description of related technology:

Various processes and reaction systems have been proposed in the past to produce tetrahydrofuran, γ-butyrolactone and 1,4-butanediol by catalytic hydrogenation of maleic acid, maleic anhydride, succinic acid and/or related hydrogenatable precursors. Various hydrogenation catalysts have also been proposed over time for this purpose, including various transition metals and combinations of transition metals supported on various inert supports, all of which are well known in the art. For example, U.S. Patent Nos. 4,985,572 and 5,149,680 disclose and claim a process for hydrogenating a carboxylic acid or anhydride thereof to the corresponding alcohol and/or carboxylic acid ester using a catalyst composition comprising an alloy of a Group VIII noble metal and another metal. In a comparative test, ruthenium and rhenium supported on a turbostratic graphite are used as catalyst in the single-pass, plug-flow hydrogenation of acetic acid. Similarly, Japanese Published Patent Applications (Kokai) 6-157490, 6-179667 and 6-157491 disclose processes for producing tetrahydrofuran by catalytic hydrogenation of maleic anhydride or maleic acid under milder conditions in the presence of an acidic substance, using as the catalyst a rhenium compound and a Group VIII metal in the first two references and a ruthenium compound in the latter. Comparative examples in the first two references illustrated the use of a catalyst of ruthenium and rhenium on carbon without the presence of the acidic substance. In general, however, these proposed catalytic reactions are predominantly carried out in an organic solvent or an organic reaction medium rather than in an aqueous solution phase. In fact, in at least one of the publications to date, water and succinic acid are considered to be detrimental to the desired catalysis, see Bulletin of Japan Petroleum Institute, Volume 12, pages 89 to 96 (1970). A notable exception to non-aqueous phase catalytic hydrogenation is the use of a carbon-supported catalyst containing from 0.5% to 10% palladium and about 1% to 10% rhenium, based on the total weight of the supported catalyst, in an aqueous solution hydrogenation reaction, the palladium on the carbon support being in the form of crystallites having an average size of about 10 nm to 25 nm and the rhenium being in a highly dispersed phase having an average size of less than about 2.5 nm, as in U.S. Patent Nos. 4,550,185; 4,609,636 and 4,659,686. Also, in a recently disclosed Japanese Kokai, September 24, 1993, the use of reduced ruthenium and tin on an activated carbon support in aqueous phase catalytic hydrogenation is disclosed. This reference teaches the use of any Group VIII noble metal, including palladium and ruthenium, in combination with either tin, rhenium or germanium. The reference distinguishes the claimed subject matter from the earlier prior art Pd, Re/carbon aqueous phase hydrogenation system by explicitly claims the use of a porous carbon support having a BET surface area of at least 2000 m²/g; a concept and limitation not characteristic of the Pd, Re/carbon system of patents 4,550,185 and 4,609,636, and a limitation which is not critical to this invention.

SUMMARY OF THE INVENTION

In view of the state of the art and in particular the limited possibilities associated with aqueous phase catalytic hydrogenation, the invention provides a highly active and robust hydrogenation catalyst system which exhibits high efficiency during the hydrogenation of a hydrogenatable precursor in an aqueous solution. More specifically, the invention provides a catalyst of ruthenium and rhenium on a carbon support, wherein both reduced components exhibit a very high degree of dispersion and a consistent compositional ratio when quantitatively analyzed at a spatial resolution in the nanometer range; i.e., any differences in their microstructure are in the sub-nanometer range. This dispersion of the reduced metals, which is evidently in the atomic or near-atomic range, results in hydrogenation rates not previously achieved with other known aqueous phase catalyst systems. The reduced ruthenium and rhenium carbon supported catalyst of the invention also has an improved lifetime, maintaining unexpectedly high hydrogenation rates over a longer period of time with a more favorable loss of catalyst activity as a function of time.

Thus, the invention provides a process for the catalytic hydrogenation of a hydrogenatable precursor in an aqueous solution, comprising the following steps:

(a) hydrogenating a hydrogenatable precursor in an aqueous solution in the presence of hydrogen and a catalyst, wherein the catalyst consists essentially of at least 0.1% by weight of ruthenium and at least 0.1% by weight of rhenium on a carbon support and these percentages are based on the total weight of the supported catalyst, wherein both ruthenium and rhenium are in a highly dispersed reduced state, characterized in that X-ray diffraction and high resolution transmission electron microscopy show no discernible metal crystallite structure, wherein the carbon support is characterized by a BET surface area of less than 2000 m²/g; and then

(b) Obtaining at least one hydrogenated product.

It is a primary object of the invention to provide a catalyst system which provides a reduced, highly dispersed state of both ruthenium and rhenium when supported on a porous carbon support, which is capable of sustaining the hydrogenation of a hydrogenatable precursor in an aqueous solution at space-time yields in excess of 600 g/kg·h. It is a further object of the invention to provide a catalyst system of reduced ruthenium and rhenium on a porous carbon support which exhibits both a high degree of dispersion and a substantially constant compositional ratio of ruthenium to rhenium when analyzed at a spatial resolution in the nanometer range across the catalyst particles and from particle to particle, indicating microstructural homogeneity in the nanometer or sub-nanometer range. The achievement of these objects, as well as other objects and the means of achieving them, will become apparent upon reading the entire specification and the appended claims.

DESCRIPTION OF THE PREFERRED EMBODIMENTS

The hydrogenation catalyst of the invention includes both ruthenium and rhenium in reduced, highly dispersed form on a porous carbon support. The catalyst comprises at least 0.1 weight percent of each metal based on the total weight of the catalyst in order to provide at least a reasonable to ensure catalyst activity. Preferably, the metal loadings on the carbon support are 0.5% to 10.0% ruthenium and 1.0% to 20.0% rhenium. A preferred catalyst for the hydrogenation of maleic acid to THF contains about 1.0% ruthenium and about 4.0% rhenium.

The carbon support useful in the invention may generally be any of the materials well known and commercially available for use in the art. Preferably, the carbon support of the catalyst is porous solid particles characterized by a size distribution typically in the range of 5 to 100 micrometers and a BET surface area typically ranging from a few 100 to nearly 2000 m2/g. Preferably, the carbon support is a commercially available material having an average particle size in the range of 20 micrometers and a BET surface area of about 900 to about 1500 m2/g. The catalyst support may be prepared to have a latently acidic, neutral or basic pH. Optionally, the catalyst support may be treated prior to application of the metal by one or more methods known in the art, such as impregnation with alkali metal salts and/or calcination.

Three different methods for preparing the highly dispersed, reduced ruthenium and rhenium on a carbon support are explained later. Each of these methods produces a hydrogenation catalyst with exceptional space-time yields when used to hydrogenate maleic acid in an aqueous solution, as also explained and illustrated in more detail later.

One of these methods consists in preparing an aqueous solution of a soluble ruthenium compound and a soluble rhenium compound and then adding this solution to the carbon support. The water is evaporated, thereby depositing the ruthenium and rhenium on the carbon support. The actual method used for adding the solution to the support may be based on any known technique, including, for example, but non-limiting: dipping, spraying, or the like. The dry or partially dried composite material is then exposed to a reducing atmosphere at elevated temperatures for a time sufficient to reduce both the ruthenium and rhenium. The dry or partially dried catalyst can then be added to the reaction zone for use as a hydrogenation catalyst.

A second method, related to the above, is to carry out the operation entirely in the presence of water or the aqueous solution of the hydrogenatable precursor. In this method, the aqueous solutions of the ruthenium compound and the rhenium compound are mixed together with the solid phase of carbon particles while the chemical reduction is carried out. The resulting catalyst exhibits the ability to sustain the hydrogenation of maleic acid in an aqueous solution at 250°C at 2000 psig total pressure at STY values in excess of 600 g/kg·h. This approach is of particular value and commercial interest because the co-deposition and co-reduction steps can actually be carried out in situ in the hydrogenation reactor and effected in the presence of the reactants, such as succinic acid and/or maleic acid. Likewise, the technique of both simultaneous deposition and simultaneous reduction serves to distinguish the ruthenium, rhenium on carbon catalyst from the palladium, rhenium on carbon catalyst taught heretofore, where sequential deposition and reduction of palladium and rhenium is required to give good hydrogenation catalyst activity. The mere fact that in the present case these simultaneous deposition and reduction techniques give the most uniform microstructure (as will be discussed and explained later in describing the importance of nanometer and sub-nanometer structure) supports the conclusion that the observed unexpectedly high catalyst activity of the Ru, Re/O catalyst (i.e. peak STY 1200 and maintained STY > 600) compared to the activity of the prior art Pd, Re/C catalyst (i.e. STY typically 300 to below 500) with its known non-uniformity and variability of Crystallite distribution and microstructure are somehow causally related or at least are a distinguishing feature. In other words, there is a fundamental difference both in the microstructure of the present catalyst system compared to the previously known aqueous hydrogenation catalyst system based on palladium and rhenium and in the unexpected increase in catalyst activity.

The third method of preparing the catalyst is to sequentially support and reduce the ruthenium on the carbon support and then add the rhenium compound so as to support and reduce the rhenium on the same carbon support. In this way, the ruthenium is pre-reduced before the rhenium is supported and then reduced. Again, as with the above co-support and co-reduction methods, the intermediate ruthenium on carbon can be achieved either in aqueous solution or in the dry state. This method also offers the possibility of carrying out the final support and reduction of the rhenium compound in situ in the hydrogenation reactor by adding the ruthenium on carbon directly to the hydrogenation reactor together with perrhenic acid or the like.

It will be understood that various other methods or alternative ways of depositing the ruthenium and/or rhenium compounds on the carbon support, all of which are known in the art, for example selective precipitation or the like, optionally with or without solvent washing to remove less desirable accompanying anions, may be considered equivalent procedures for preparing the catalyst of the invention.

The ruthenium compounds which can be used in the invention to prepare the catalyst can generally be any of the compounds which are either water-soluble or can be easily converted into a water-soluble or partially water-soluble compound which can be applied to the carbon support. However, this would not, for example, limitingly include such compounds as RuCl₃·xH₂O, Ru(NO)(NO₃)₃ and the like. Preferably, ruthenium trichloride is used.

The rhenium compounds useful in the invention to prepare the catalyst can generally be any of the compounds that are either water soluble or that can be readily converted to a water soluble or partially water soluble rhenium compound by the action of an oxidizing acid solution, hydrogen peroxide or the like. As such, aqueous perrhenium chemistry including the application of the corresponding rhenium oxide is particularly useful. Preferably, a perrhenic acid solution is used.

The reducing agent used to chemically reduce the ruthenium and/or rhenium can generally be any reducing agent or reducing environment compatible with either liquid phase reduction or vapor phase reduction, for example but not limited to: formaldehyde, hydrazine hydrate, hydroxylamine, sodium hypophosphite, sodium formate, glucose, acetaldehyde, sodium borohydride, hydrogen, and the like. When vapor phase reduction, including gaseous hydrogen with or without an inert diluent gas such as nitrogen or the like, is employed in the presence of the solid catalyst precursor, the vapor phase reduction is typically carried out at a temperature in the range of 100 to 500°C (preferably 250-300°C) at atmospheric pressure.

The process using the highly dispersed reduced ruthenium and rhenium on carbon catalyst to hydrogenate a hydrogenatable precursor according to the invention can be carried out by various modes of operation as generally known in the art. Thus, the general hydrogenation process can be carried out by using a fixed bed reactor, various types of stirred slurry reactors or the like, either on batch or continuous manner, wherein an aqueous liquid phase containing the hydrogenatable precursor is in contact with a gaseous phase containing hydrogen at elevated pressure and the solid catalyst in particulate form. Typically, such hydrogenation reactions are carried out at temperatures of from about 100°C to about 300°C in sealed reactors maintained at pressures of from about 1000 to about 3000 psig.

For the purposes of the invention, the hydrogenatable precursor can be broadly defined as any compound or material that can be chemically reduced by hydrogenation or hydrogen uptake. This would particularly include, but again not by limitation, various organic compounds containing unsaturated groups or oxidized organic functional groups, or both. In particular, the catalytic hydrogenation of maleic acid to γ-butyrolactone and tetrahydrofuran in the aqueous phase illustrates the utility of the process of the invention. In this context, and as illustrated in the examples, it should be understood that various products of the sequential hydrogenation reaction are also potential hydrogenatable precursors. In other words, it is known that in the conversion of maleic acid to tetrahydrofuran, the chemical reduction is sequential, involving the rapid addition of hydrogen to the double bond, converting maleic acid to succinic acid, followed by the slower addition of hydrogen, forming potential intermediates such as γ-butyrolactone and/or butanediol and finally the desired tetrahydrofuran (which corresponds to the uptake of five moles of H and the formation of three moles of H₂O per mole of THF). In industrial production, the overall selectivity towards THF production can be significantly influenced by the optimization of the reaction conditions, including the maintenance of an appropriate degree of acidity, thus favoring ring closure and the formation of cyclic ether at the expense of the formation of diol, the continuous removal of the vapor from the more volatile products and the subsequent separation and recycling of the lactone. It is clear that in this case the γ-Butyrolactone can be considered either as a byproduct or as a recycled hydrogenatable precursor reactant.

For the purposes of the invention, when reference is made to highly dispersed reduced phases having a substantially constant compositional ratio of ruthenium to rhenium when analyzed with a spatial sensitivity on the order of one square nanometer, reference is made to the quantitative analysis of the ruthenium and rhenium elements observed or measured in a region of the catalyst support down to physical dimensions on the order of one nanometer. This expression describes a degree of homogeneity present at relatively small physical dimensions, which is consistent with the previous statement that all microstructural differences associated with the reduced metals are in the nanometer or sub-nanometer range. This in turn is consistent with the notion that in the catalyst of the invention the respective metals present on the catalyst support are present as a dispersion on the order of atoms or nearly atoms. However, it should also be understood that this statement and these observations and measurements do not exclude the presence of additional microstructures with dimensions exceeding the nanometer limit, nor observable changes in the microstructure slightly above the nanometer range related to the choice of manufacturing processes and to fatigue of the catalyst during use, nor observable labile structural features below the nanometer observation limit (all of which are currently believed to be characteristic to some extent of the present catalyst system).

For example, and conversely, it is well known that the preparation of a bimetallic catalyst containing rhenium (application of rhenate as rhenium heptaoxide) explicitly involves mobility (see SB Ziemecki., GA Jones and JB Michel, "Surface Mobility of Re₂O₇ in the System Re⁷+Pd⁰/γ-Al⁷O₃ ", Journal of Catalysis 99, pp. 207-216, 1986), thereby at least allowing the possibility of migration of rhenium is assumed to be below the limit of perception. With respect to catalyst fatigue during use, a perceptible ruthenium phase has been observed in used catalysts where initially no perceptible ruthenium or rhenium was present, suggesting some form of mobility associated with the ruthenium. However, the evolution of perceptible ruthenium during use is generally within the range of the average pore size of the porous carbon support (i.e., typically 20 Å or so), which is still a more dispersed state than that previously mentioned in connection with the prior art catalyst systems such as palladium and rhenium on carbon. Perceptible ruthenium may also be observed depending on the method of preparation for the catalyst. Thus, a catalyst prepared by sequential deposition and reduction, which involves depositing and reducing the ruthenium before depositing and reducing the rhenium, can exhibit some discernible ruthenium microstructure of a few nanometers or slightly larger. The deposition and reduction of ruthenium in the absence of rhenium can produce discernible ruthenium arrangements that are unrelated to the microporous structure of the carbon and that do not necessarily interfere with the development of the desired highly active catalyst, which is believed to be related to the highly dispersed state of the metals at a constant composition ratio. Typically, the co-deposited and subsequently co-reduced catalyst is remarkably free of any discernible phase or surface crystallite microstructure and as such is considered the preferred catalyst for the purposes of the invention.

The following examples are provided to demonstrate and further clarify various individual aspects and features of the invention in more detail, while the comparative examples and data are intended to further illustrate the differences and advantages of the invention. The examples are not intended to be limiting and do not illustrate the invention in any way. unduly limit, particularly with regard to the ultimate properties of the improved catalyst and the utility of the claimed improved process.

EXAMPLE 1

Aqueous solutions of RuCl3.xH2O (2.0305 grams of a 1% ruthenium solution) and HReO4 (1.104 grams of a 7.7% rhenium solution) were mixed and the resulting solution was added to 2.03 grams of a particulate carbon support (average particle size about 20 µm) which was itself acidic (pH = 4 - 4.5) with a BET surface area of about 1500 m2/g. The mixture was dried overnight at 110°C under reduced pressure (in vacuo) under a nitrogen purge. The dried catalyst was then placed in 150 cc quartz boats in a 1500 cc quartz tube with a 6.35 cm (2.5 inch) diameter, horizontally oriented in a tube furnace. A nitrogen purge was carried out for 15 minutes at room temperature at a flow of 500 cc/min. The flow was switched to helium at 100 cc/min and held for 25 minutes. The temperature was then increased to 150°C for one hour. After the one hour hold, 100 cc/min of hydrogen was added to the flow and it was held at 150°C for one hour. The temperature was then increased to 300°C with a 1:1 H2-He flow and then held for eight hours. After this period, the flow was switched to helium and the catalyst was cooled in helium overnight. Once the temperature was below 50°C, the catalyst was passivated in flowing 1% O2 in N2 for 30 minutes. The resulting catalyst of 1 wt.% Ru and 4 wt.% Re on carbon was tested in the reactor as described in the following examples. When examined by conventional X-ray diffraction and high resolution transmission electron microscopy, the catalyst showed no discernible metal crystallite structure (i.e. no separate structures with a size > 10 Å apart from the background carbon structure). However, chemical analysis by EDX confirms that the weight ratio of one to four of the ruthenium to the rhenium on the carbon particle and from particle to particle was essentially uniform.

Thus, it should be understood that for the purposes of the claims of the invention, the statement "wherein both ruthenium and rhenium are dispersed at a constant composition ratio of ruthenium to rhenium when STEM analyzes the catalyst particles with a spatial resolution on the order of one square nanometer" has the meaning illustrated above. The measurement was made using field emission STEM, which uses a highly collimated electron beam incident on approximately one square nanometer of the catalyst surface (i.e., passing the electron beam through the catalyst with a cross-sectional diameter of one square nanometer).

EXAMPLE 2

2.0216 grams of an aqueous ruthenium trichloride solution containing 1% ruthenium was added to 2.07 grams of a particulate carbon support having a BET surface area of 1500 m2/g and being acidic itself (pH 4 - 4.5). The mixture was dried overnight at 110°C under reduced pressure (in vacuo) with a nitrogen purge. Reduction of the ruthenium metal salt supported on the carbon support was then carried out on the dry material essentially as described in Example 1.

EXAMPLE 3

0.5100 grams of a 1% ruthenium solution derived from RuCl3.xH2O was mixed with 0.2780 grams of a HReO4 solution (7.7% Re from Re-07). To this was added 35.2574 grams of a 7% succinic acid solution. This solution was placed in a 125 ml Hastelloy C autoclave along with 0.503 grams of an acidic particulate carbon (B ET 1.500 m2/g). The reactor was checked for leaks with nitrogen and then hydrogen was added to the reactor. and the reaction was carried out as described in Examples 4-15.

EXAMPLES 4 to 15

To evaluate the ruthenium/rhenium on carbon catalysts of the invention, a series of batch hydrogenation reactions were carried out using succinic acid as a hydrogenatable precursor (i.e., reactant). The equipment used was a 125 mL autoclave purchased from Autoclave Engineers of Erie, Pennsylvania. The autoclave was equipped with a magnetically coupled stirrer driven by an air motor. A set of removable baffles (three vertical strips of Hastalloy C welded to two Hastalloy C rings) were used to assist in stirring. Gases were pumped into the reactor using a Haskell pump capable of pumping up to 17.24 MPag (2500 psig). The nitrogen used was CP grade from a gas cylinder, the nitrogen came from a gas cylinder, the succinic acid was Baker Analyzed Reagent Grade Crystal, and the water was deionized water. 7% by weight solutions of succinic acid were prepared in 500 gram batches by dissolving 35 grams of succinic acid in 465 grams of deionized water. The solutions were heated to 30ºC to completely dissolve the succinic acid. The solutions were then titrated to accurately determine the percent acid. The procedure used involved adding the catalyst (typically about 0.1 to 0.5 grams) to the reactor, followed by 35 grams of 7% succinic acid solution. The header was then placed on the reactor, screwed into place, and rotated downward. Nitrogen was then pumped into the reactor using the Hastell pump to 2400 psiag to check for any leaks in the system. If the pressure did not remain constant, leaks were detected and repaired using Snoop. If no leaks were detected, the nitrogen was slowly vented. After the nitrogen was vented to atmospheric pressure, hydrogen was pumped into the reactor using the Haskell pump to 8.27 Mpag (1200 psig). The stirrer was set at 700 rpm and the heater was turned on. The contents of the reactor were heated to 250ºC, which took approximately 1 hour. When the internal temperature reached 250ºC, stirring was continued for 3 hours. After stirring at 250ºC for 3 hours, the heater was turned off and cooling water was introduced into the jacket surrounding the reactor. When the reactor contents had cooled to below 25ºC (approximately 0.5 hour), the air motor was turned off and the hydrogen was vented very slowly. Once the hydrogen had been vented to atmospheric pressure, the screws were removed. The contents were quickly withdrawn from the reactor using a 60 mL disposable syringe with a large bore stainless steel needle. After the sample was removed from the reactor, 10 mL was filtered using a 10 mL disposable syringe and a 0.45 micron PTFE syringe filter. A sample of this was analyzed by GC for tetrahydrofuran, 1-propanol, n-butanol, gamma-butyrolactone, and 1,4-butanediol using a Hewlett-Packard 5890A instrument, and a portion was titrated for residual succinic acid. For titration, approximately 8 grams of filtered solution was weighed into a 150 mL beaker, recording the weight to 4 decimal places. A magnetic stir bar and stir plate were used to stir the solution. A few drops of phenolphthalein solution were added, and the solution that splashed on was rinsed from the rim with deionized water. Titration with 0.1 N NaOH solution was carried out to the pink endpoint (30 seconds). The resulting data for various catalysts prepared according to the invention, as well as various controls characteristic of previously known aqueous hydrogenation catalysts of palladium and rhenium on a carbon support, are given in the following table. TABEL

** all data in mmol, with THF including the vapor phase, based on a carbon mass balance calculation

(a) Ru and Re co-reduced in H2: aqueous solution of RuCl3 and HReO4 deposited on carbon, dried overnight at 110°C in vacuo with N2 purge, then co-reduced for 8 hours at 300°C in a 1:1 mixture of H2/He, see Example 1.

(b) Ru pre-reduced in H2: 1% Ru was deposited on carbon as RuCl2 solution, dried overnight at 110°C in vacuo with N2 purge, then reduced for 8 h at 300°C in a 1:1 mixture of H2/He and added to the reactor with HReO4 solution, see Example 2.

(c) Ru and Re reduced together in the reactor: aqueous solution of RuCl₃ and HReO₄ with 35 g of 7% succinic acid with particulate carbon added to the reactor, purged with N₂ and then with 8.27 MPa (1200 psi) H₂ at 250°C, final pressure 15.17-15.86 Mpag (2200-2300 psiag) for in situ co-reduction, see Example 3.

Carbon A: particulate carbon with a BET surface area of 1500 m²/g and intrinsically acidic (pH = 4 - 4.5)

Carbon B: particulate carbon with a BET surface area of 1000 m/g and intrinsically basic (pH = 8 - 9)

EXAMPLE 16

Continuous runs were carried out in a backmixed slurry reactor by adding the catalyst prepared in 150 ml of water according to Example 1 to a 300 ml Hastelloy C autoclave equipped with a stirrer, a thermocouple, inlets for hydrogen and maleic acid, and an outlet through which the product was removed with the excess hydrogen and water vapor. Prior to reaction, the catalyst slurry was activated for one hour at 250°C under a hydrogen flow of 1000 ml/min at 2000 psig. Maleic acid was then added as a 40 wt.% aqueous solution at various feed rates and the reactor was maintained at 2000 psig and 250°C. The volatile products and water were removed from the reactor at a rate determined by the hydrogen feed rate. The hydrogen feed rate was adjusted to balance the amount of water carried out with the exiting hydrogen gas with the amount of water supplied with the maleic acid feed and the amount generated by the reaction. The reactor level was kept constant at 100-200 cc. In all cases, an excess of hydrogen was added relative to the amount consumed by the reaction and it was found that the hydrogen feed rate did not affect the catalyst performance.

The maleic acid solution was added at a low initial feed rate, which was increased by 2-3 cm³/h approximately every 8 hours until the acidity of the reactor reached 8%. Thereafter, the maleic acid feed rate was adjusted as required to maintain the reactor acidity between 6-10%.

The runs were carried out continuously for several weeks and the periods during the steady state operation were selected for analysis, generally ranging from 8 to 24 hours. The product composition data obtained during the steady state operation were averaged to obtain the average production rates (g/h) of THF (tetrahydrofuran), BDO (1,4-butanediol), GBL (gamma-butyrolactone), PrOH (n-propyl alcohol), BuOH (n-butyl alcohol) and alkanes (primarily butane and methane). The product composition was measured by condensing a portion of the volatile products and water in the exit gas stream and collecting the liquid product. The volume of liquid product collected every hour was measured and its composition was analyzed using a calibrated gas chromatograph (GC) equipped with a flame ionization detector. The remaining non-condensed products (THF and alkanes) still remaining in the exit gas stream were analyzed by measuring the gas flow rate and then analyzing the composition of the gas stream every two hours using GC procedures similar to those used for analyzing the liquids. The reactor contents were sampled every four hours and analyzed by GC and titration. A titration with sodium hydroxide was used to monitor the acid concentration in the reactor and the results are reported as weight percent succinic acid. GC analysis was performed using a Supelcowax-10 capillary column (30 m x 0.052 mm) which was maintained at 75ºC for 5 minutes after injection and then heated at 10ºC per minute to a final temperature of 200ºC. The combination of these three analytical methods allows the calculation of catalyst performance in terms of space-time yield (STY) (calculated as grams of THF produced per kilogram of catalyst per hour), product selectivity and mass balance for each run.

In the first run, 3.5 grams of catalyst (dry weight) containing 1% ruthenium and 4% rhenium on a carbon support was used. The initial maleic acid feed rate was 6 cm3/h and this was increased in 2 cm3/h increments up to 20 cm3/h at which point the acidity in the reactor reached 8%. Thereafter, the coating rate was adjusted to maintain the acidity at 6-10%. The initial THF production rate (at the lowest maleic acid feed rate) was 200 STY (g THF/kg catalyst·h). As the feed rate was increased, the production rate continued to increase, reaching STY up to 2200 once the feed rate reached 20 cm3/h. During the next three weeks of operation, activity gradually decreased, eventually reaching a THF production rate of about 600 RZA.

A second comparative run was conducted as the above run, except that the catalyst was replaced with 10 g of the same carbon support containing 1% palladium and 4% rhenium. The maleic acid feed started at 12 cc/h and was increased to 28 cc/h over five days. During this period, the reactor acidity increased up to 5% and the THF production rate increased to an STY maximum of 495.

It should be understood that for the purposes of the claims of the invention, the statement "the catalyst is characterized by a space-time yield of conversion of maleic acid to tetrahydrofuran of over 600 grams of product per kilogram of catalyst per hour at 250°C and 13.79 MPa (200 psig) total pressure" means a catalyst which behaves substantially as illustrated in Example 16. In other words, the statement is intended to describe the highly active and robust hydrogenation catalyst system of the invention in terms of the ability of the catalyst to sustain THF production in a continuous operation with the volatile product being removed by the excess gas phase containing the hydrogen reactant under all of the specified conditions (i.e., 250°C and 13.79 MPa (2000 psig) total pressure). However, it is even more important that the reaction described be carried out such that the catalyst can exhibit its full reaction potential and is not subject to stoichiometric limitations of the particular reactants (i.e., the hydrogen or the hydrogenatable precursor). For pragmatic reasons, and as illustrated in the example, the feed rate of reactants to the reactor is adjusted (increased) until the catalyst is fully utilized. Typically, this is accomplished by controlling the residual acidity of the reactor and increasing the reactant flow until a steady state of at least about 5% acid or more exists in the reactor. At acid concentrations below this value, STY is not optimized and the tendency for overhydrogenation of the product also reduces THF selectivity. Experience shows that operation at or above this residual acidity gives STY values that have not been achieved with the prior art catalyst systems. It was possible to achieve residual acid levels in the reactor of up to twenty percent, but it should be remembered that succinic acid solutions of this concentration solidify if they are not kept at high temperatures.

The advantages and improvements of the invention are significant and numerous. First and foremost, the catalyst system of the invention provides a process for the sustained catalytic hydrogenation of maleic acid dissolved in water with previously unattained space-time yields. Furthermore, this hydrogenation in the aqueous phase is not achieved by the use of external additives or aids such as the presence of strong acids, nor does the hydrogenation depend on the use of carbon supports with extremely high BET surface areas. Therefore, the catalyst and process are suitable for continuous industrial operation involving the addition of maleic acid to the catalytic reactor as an aqueous solution and the continuous recovery of the gaseous product by removal of the other reactant and hydrogen from the reaction zone, all of which represents and creates a significant economic advantage, particularly with respect to the industrial scale of current THF production. Further cost savings are provided by the inherently lower cost of the respective metals of the catalyst and the lower investment requirement per unit of production related to the high STY of the catalyst. Likewise, the discovery of the high dispersion in the nanometer or sub-nanometer range enables a method to characterize the catalyst and its respective performance, which in turn offers a planning advantage in the operation of an industrial plant.

Claims (7)

1. A process for the catalytic hydrogenation of a hydrogenatable precursor in an aqueous solution, comprising the steps of: (a) hydrogenating a hydrogenatable precursor in an aqueous solution in the presence of hydrogen and a catalyst, the catalyst consisting essentially of at least 0.1 wt.% ruthenium and at least 0.1 wt.% rhenium on a carbon support, the percentages being based on the total weight of the supported catalyst, both ruthenium and rhenium being in a highly dispersed reduced state characterized by X-ray diffraction and high resolution transmission electron microscopy showing no discernible metal crystallite structure, the carbon support being characterized by a BET surface area of less than 2000 m²/g; and then (b) obtaining at least one hydrogenated product. 2. A process according to claim 1, wherein the hydrogenatable precursor is selected from the group consisting of maleic acid, maleic anhydride, fumaric acid, succinic acid, maleic acid, the esters corresponding to these acids, γ-butyrolactone and mixtures thereof. 3. A method according to claim 1, wherein both the ruthenium and the rhenium are dispersed in the sub-nanometer range. 4. A process according to claim 1, wherein both the ruthenium and the rhenium are highly dispersed with a constant composition ratio of ruthenium to rhenium when measured by STEM on the catalyst particles analyzed with a spatial resolution on the order of one square nanometer. 5. A process according to claim 1, wherein the catalyst is prepared by the process comprising, in this sequence, the steps of: supporting both a ruthenium compound and a rhenium compound on a carbon support; drying the carbon support containing both the ruthenium compound and the rhenium compound; and then reducing the ruthenium and the rhenium. 6. A process according to claim 1, wherein the catalyst is prepared by the process comprising, in this sequence, the steps of: supporting both a ruthenium compound and a rhenium compound on a carbon support and then reducing the ruthenium and the rhenium in situ in a hydrogenation reactor, wherein the hydrogenation of the hydrogenatable precursor takes place in an aqueous solution. 7. A process according to claim 1, wherein the catalyst is prepared by the process comprising, in this sequence, the steps of: applying a ruthenium compound to a carbon support; drying the carbon support containing the ruthenium compound, and then reducing the ruthenium before applying a rhenium compound to the carbon support and reducing it.